Article
pubs.acs.org/JPCA
Rotational Spectrum and Conformational Composition of
Cyanoacetaldehyde, a Compound of Potential Prebiotic and
Astrochemical Interest
Harald Møllendal,*,† Laurent Margulès,‡ Roman A. Motiyenko,‡ Niels Wessel Larsen,§
and Jean-Claude Guillemin∥
†
Centre for Theoretical and Computational Chemistry (CTCC), Department of Chemistry, University of Oslo, P.O. Box 1033
Blindern, NO 0315 Oslo, Norway
‡
Laboratoire de Physique des Lasers, Atomes, et Molécules, UMR CNRS 8523, Université de Lille I, F-59655 Villeneuve d’Ascq
Cédex, France
§
Department of Chemistry, University of Copenhagen, The H. C. Ørsted Institute, Universitetsparken 5, DK 2100 Copenhagen Ø,
Denmark
∥
Sciences Chimiques de Rennes, École Nationale Supérieure de Chimie de Rennes, CNRS, UMR 6226, Avenue du Général Leclerc,
CS 50837, 35708 Rennes Cedex 7, France
S Supporting Information
*
ABSTRACT: The rotational spectrum of cyanoacetaldehyde
(NCCH2CHO) has been investigated in the 19.5−80.5 and
150−500 GHz spectral regions. It is found that cyanoacetaldehyde is strongly preferred over its tautomer cyanovinylalcohol (NCCHCHOH) in the gas phase. The spectra of
two rotameric forms of cyanoacetaldehyde produced by
rotation about the central C−C bond have been assigned.
The C−C−C−O dihedral angle has an unusual value of
151(3)° from the synperiplanar (0°) position in one of the
conformers denoted I, while this dihedral angle is exactly
synperiplanar in the second rotamer called II, which therefore
has Cs symmetry. Conformer I is found to be preferred over II by 2.9(8) kJ/mol from relative intensity measurements. A double
minimum potential for rotation about the central C−C bond with a small barrier maximum at the exact antiperiplanar (180°)
position leads to Coriolis perturbations in the rotational spectrum of conformer I. Selected transitions of I were fitted to a
Hamiltonian allowing for this sort of interaction, and the separation between the two lowest vibrational states was determined to
be 58794(14) MHz [1.96112(5) cm−1]. Attempts to include additional transitions in the fits using this Hamiltonian failed, and it
is concluded that it lacks interaction terms to account satisfactorily for all the observed transitions. The situation was different for
II. More than 2000 transitions were assigned and fitted to the usual Watson Hamiltonian, which allowed very accurate values to
be determined not only for the rotational constants, but for many centrifugal distortion constants as well. Two vibrationally
excited states were also assigned for this form. Theoretical calculations were performed at the B3LYP, MP2, and CCSD levels of
theory using large basis sets to augment the experimental work. The predictions of these calculations turned out to be in good
agreement with most experimental results.
■
INTRODUCTION
The C3−C1−C2−O4 dihedral angle can conveniently be
used to define the conformational properties of cyanoacetaldehyde. Two rotameric forms may exist for this compound. The
said dihedral angle turned out in the course of this work to be
roughly 150° in I, halfway between anticlinal (120°; obsolete:
gauche) and antiperiplanar (180°; obsolete: trans). This
dihedral angle is 0° in II, which therefore takes a synperiplanar
(obsolete: cis) conformation.
The rotational spectrum of cyanoacetaldehyde (NCCH2CHO)
has not previously been investigated. There are several reasons
why the present investigation was undertaken. First, cyanoacetaldehyde may exist in the gas phase as a mixture of rotameric
forms. Moreover, its tautomer, 3-hydroxy-2-propenenitrile
(cyanovinylalcohol; HOCHCHCN), exists in equilibrium
with cyanoacetaldehyde. In Figure 1, models of two conformers
(denoted I and II) of cyanoacetaldehyde and three forms (III−
V) of the tautomer cyanovinylalcohol are shown with atom
numbering indicated on I and III.
© 2012 American Chemical Society
Received: December 21, 2011
Revised: March 20, 2012
Published: March 20, 2012
4047
dx.doi.org/10.1021/jp212306z | J. Phys. Chem. A 2012, 116, 4047−4056
The Journal of Physical Chemistry A
Article
who already in 1903 reported a procedure for the opening of
isoxazole by sodium ethylate followed by in situ trapping of the
formed salt of cyanovinylalcohol.3 The first isolation of
cyanoacetaldehyde in an analytical scale was achieved in 1991
by flash vacuum thermolysis of a Meldrum acid derivative.4 The
mass spectrum of a crude mixture was obtained the following
year.5 Neat cyanoacetaldehyde, in sufficiently pure and in large
enough quantities to allow a full spectroscopic study, was first
prepared as recently as in 2007 by flash vacuum pyrolysis of
isoxazole at 770 °C and isolated from impurities using a cold
trap at −42 °C after the pyrolysis oven.1 Cyanoacetaldehyde
was found to be a moderately kinetically stable compound at
room temperature and much more reactive than acetaldehyde.1
Subsequent characterization by 1H NMR spectroscopy of
solutions of cyanoacetaldehyde showed the presence of both
cyanoacetaldehyde and cyanovinylalcohol. The concentration
of each of these tautomers was found to be strongly dependent
on the solvent.1
It is interesting to note that the presence of a nitrile group in
cyanoacetaldehyde makes the aldehyde form significantly more
stable than the enol form, while the opposite is the case for the
corresponding thiol and amine, which are vinyl derivatives,
namely, Z- and E-HSCHCHCN,6,7 and Z- and E-H2NCH
CHCN,8−10 respectively. Microwave (MW) spectra of the
Z-thiol7 and the Z-amine10 have already been reported.
Rotational spectra have been used to identify the vast
majority of molecules found in the interstellar medium.11,12
Prebiotic existence of cyanoacetaldehyde is quite likely because
of the facile hydrolysis of cyanoacetylene,2 a compound that is
formed in a spark discharge reaction in a methane−nitrogen
mixture,13 is prevalent in the interstellar medium,11 and found
in comets14 and in Titan’s atmosphere.15 A gas-phase model of
an uncatalyzed addition reaction between the two neutral
molecules cyanoacetylene and water has been explored using
quantum chemical calculations and found to be inefficient.16
Protonated cyanoacetylene (H−CC−CN−H+) also exists
in the interstellar medium.17 The addition of water to this
cation in the gas phase followed by an electron recombination
reaction was also predicted to be inefficient but to a much lesser
extent than the hydrolysis of cyanoacetylene.16 The formation
of cyanoacetaldehyde in liquid water, nevertheless, occurs,2
presumably formed in a catalytic process.16
All this indicates that cyanoacetaldehyde may exist in the
interstellar space or in planetary atmospheres. The title
compound consists of eight atoms. Modern radio astronomy
is now capable of detecting interstellar compounds of this
complexity by means of their rotational spectra.11,12 The
rotational spectrum of cyanoacetaldehyde presented herein
could therefore be the key for its identification anywhere in the
Universe.
There is another important reason for performing a study of
the rotational spectrum of cyanoacetaldehyde. The 1979
discovery of pyrimidines in the carbonaceous meteorites18
showed that prebiotic synthesis of these important biomolecules, which are constituents of RNA and DNA, indeed occurs.
Cyanoacetaldehyde has been demonstrated to react under
various conditions with a number of likely prebiotic molecules
to form pyrimidines, even in high yields.2,19−23 This interesting
potential prebiotic connection to pyrimidines should also be
kept in mind.
Our experimental work has been augmented by high-level
quantum chemical calculations, which were undertaken to
obtain information for use in assigning the rotational spectrum
Figure 1. Conformers I and II are rotamers of cyanoacetaldehyde
(CH2(CN)CHO), whereas III−V are cyanovinylalcohol tautomers.
Atom numbering is given on I and III. The MP2/aug-cc-pVTZ energy
differences corrected for zero-point vibrational energies are given
relative to the energy of I. Conformers I and II were found
experimentally. Rotamer I was found to be 2.9(8) kJ/mol more stable
than II by relative intensity measurements.
Cis and trans configurations exist for the tautomer
cyanovinylalcohol. The cyano and alcohol groups are in a cis
configuration in III and V, and in a trans configuration in IV.
Rotation about the C2−O4 bond may result in rotational
isomerism in each of the two tautomers. The three planar
species III−V shown in Figure 1 represent typical forms of
cyanovinylalcohol. It has been shown by NMR spectroscopy1
that cyanoacetaldehyde predominates over cyanovinylalcohol in
solution. Further forms of cyanovinylalcohol may exist, but they
are likely to have even higher energies relative to I than
conformers III−V. These very high-energy forms will
consequently have extremely weak rotational spectra, and
they have therefore not been considered further.
The synthesis of cyanoacetaldehyde is not straightforward.
Hydrolysis of cyanoacetylene (HCC−CN) in water leads
to the formation of cyanoacetaldehyde,2 but the problem is that
it has not been possible to isolate cyanoacetaldehyde from the
water solution. It is convenient, almost necessary, to use a neat
sample in order to investigate a rotational spectrum. The first
progress to isolate cyanoacetaldehyde was made by Claisen,
4048
dx.doi.org/10.1021/jp212306z | J. Phys. Chem. A 2012, 116, 4047−4056
The Journal of Physical Chemistry A
Article
Scheme 1. Synthesis of Cyanoacetaldehyde
measurement accuracy for isolated lines is estimated to be
better than 30 kHz. However, if the lines were blended or had a
poor signal-to-noise ratio, their uncertainties were estimated to
be 100 or even 200 kHz. The inverse squares of the
uncertainties were used as weights in the least-squares fit of
the spectrum. The line width was limited by Doppler
broadening. The present measurements were performed at
room temperature. The sample pressure during measurements
was about 1.5 Pa (15 μbar). The filling of the cell was
performed in the same way as in Oslo. We found that the
sample was stable up to 7 h at this low pressure.
The Lille Fourier transform microwave spectrometer coupled
to a pulsed molecular beam (MB-FTMW) was recently
updated.29 This device was used in an attempt to assign
transitions of the two rotamers assigned in this work involving
low values of the principal J quantum number. The sample was
put in a reservoir, under a neon buffer gas pressure of 1 bar, and
heated up to 50 °C. The a-type transitions of rotamer I were
searched in the 5 GHz (101−000 transition), 10 GHz (202−101),
and 15 GHz (303−202) regions. Unfortunately, no signals were
observed. This was probably due to the specific experimental
conditions (i.e., long residual time in a heated reservoir)
associated with the MB-FTMW spectrometer, which was not
compatible with the chemical stability of cyanoacetaldehyde.
Quantum Chemical Methods. The present quantum
chemical calculations were performed employing the Gaussian03 suite of programs,30 running on the Titan cluster in
Oslo. Becke’s three-parameter hybrid functional31 employing
the Lee, Yang, and Parr correlation functional (B3LYP)32 were
employed in the density functional theory (DFT) calculations.
Møller−Plesset second order perturbation theory (MP2)
calculations33 and coupled-cluster calculations with singlet
and doublet excitations (CCSD)34,35 were also undertaken. The
CCSD calculations are very costly and were speeded up by
making use of a B3LYP force field that was calculated prior to
the CCSD calculations. Peterson and Dunning’s36 correlationconsistent aug-cc-pVTZ basis set, which is of triple-ζ quality
and includes both diffuse and polarized functions, was used in
the B3LYP and MP2 calculations. The smaller basis set ccpVTZ36 was employed in the CCSD calculations. This set does
not include diffuse functions.
and investigating properties of the potential-energy hypersurface.
■
EXPERIMENTAL SECTION
Synthesis. Cyanoacetaldehyde is a kinetically unstable
compound at room temperature, and its revaporization is
really challenging. The flash vacuum thermolysis at 770 °C
under 0.1 mbar of isoxazole, a commercially available
compound, led to the formation of cyanoacetaldehyde in a
24% yield. Selective trapping at −28 °C allowed us to obtain a
pure compound (Scheme 1).1 After several unsuccessful
attempts to heat it as slowly as possible in vacuo (0.1 mbar)
to have the lowest temperature of vaporization, we found that a
more efficient way to vaporize it was a fast heating in a 50 °C
water bath followed, when the expected vapor pressure was
achieved, by a fast cooling with a liquid nitrogen bath to save
the residual product. This unusual, but efficient, experiment was
repeated up to 5 times with the same sample before the
complete decomposition of the product. 1H and 13C NMR
spectra of all the vaporized samples showed the presence of a
pure product. We estimated that up to 20% of the compound
can be revaporized under these conditions, where black
oligomeric compounds were the main decomposition products.
Spectroscopic Experiments in Oslo. The MW spectrum
of cyanoacetaldehyde was studied using the Stark-modulation
MW spectrometer of the University of Oslo operating in the
7−120 GHz spectral range. Details of the construction and
operation of this device have been given elsewhere.24−26 This
spectrometer has a resolution of about 0.5 MHz and measures
the frequency of isolated transitions with an estimated accuracy
of ∼0.10 MHz. The 19.5−80.5 GHz frequency interval was
recorded. Radio frequency microwave double-resonance experiments (RFMWDR), similar to those performed by Wodarczyk
and Wilson,27 were conducted to unambiguously assign
particular transitions, using the equipment described elsewhere.24 The sample, which was found to polymerize at room
temperature, was kept at liquid-nitrogen (−196 °C) or dry ice
temperature (−78.5 °C) when not in use. The ampule
containing the sample had to be heated quickly at 50 °C
with a water bath in order to fill the MW cell with a fresh
portion, which was studied at a pressure of about 10 Pa. The
formation of oligomeric brownish products was seen to occur
during the heating of the sample.
Spectroscopic Experiments in Lille. The submillimeterwave measurements (150−500 GHz) were performed employing the Lille spectrometer,28 which is composed exclusively of
solid-state devices as sources. The frequency of the Agilent
synthesizer (12.5−17.5 GHz) was first multiplied by six and
amplified by a Spacek active sextupler providing the output
power of +15 dBm in the W-band range (75−110 GHz). This
power is high enough to use passive Schottky multipliers (×2,
×3, and ×5) from Virginia Diodes Inc. in the next stage of the
frequency multiplication chain. The detector is an InSb liquid
He-cooled bolometer from QMC Instruments Ltd., which
improves the sensitivity of the spectrometer. The sources were
frequency modulated at 10 kHz. The absorption cell is a
stainless-steel tube (6 cm diameter, 220 cm long). The
■
RESULTS AND DISCUSSION
Quantum Chemical Calculations. MP2/aug-cc-pVTZ
calculations were first performed in order to get an overview
of which species might contribute significantly to the rotational
spectrum. The fully optimized structures, dipole moments, and
harmonic vibrational frequencies were calculated for conformers I and II of cyanoacetaldehyde and for the
cyanovinylalcohol tautomers III−V depicted in Figure 1.
Only positive vibrational frequencies were found for each
species, which is indicative of a minimum (stable conformer)
on the energy hypersurface. The results of these calculations are
summarized in Tables 1S−5S of the Supporting Information.
It is seen from these tables that conformer I of
cyanoacetaldehyde is calculated to be the global energy
minimum of these five species. The MP2 energies of the
4049
dx.doi.org/10.1021/jp212306z | J. Phys. Chem. A 2012, 116, 4047−4056
The Journal of Physical Chemistry A
Article
other forms relative to the energy of I are indicated on Figure 1.
These values have been corrected for zero-point vibrational
energies. Conformer II of cyanoacetaldehyde was found to have
a symmetry plane with two out-of-plane hydrogen atoms (Cs
symmetry). This conformer is predicted to be 2.3 kJ/mol less
stable than I, while the cyanovinylalcohol species, III, is found
to be 6.6 kJ/mol less stable and to be the lowest-energy forms
of the three cyanovinylalcohol forms shown in this figure. This
is consistent with the fact that III is stabilized by an internal
hydrogen bond between the cyano and the hydroxyl group.
Such stabilization is not possible for other forms of
cyanovinylalcohol.
Rotamers IV and V are calculated to have significantly higher
energies (∼17−20 kJ/mol) than I, and their rotational spectra
are, for this reason, not considered further because they would
presumably be extremely weak at room temperature.
The MP2 structures (Tables 1S−5S, Supporting Information) reveal no unusual structural parameters with one
important exception, namely, the C3−C1−C2−O4 dihedral
angle of conformer I. This dihedral angle is calculated to be
144.9°. This unusual angle deviates by about 25° from the
canonical 120° angle, corresponding to an anticlinal conformation.
CCSD/cc-pVTZ calculations, which are at a much higher
level of theory than MP2 calculations, were also undertaken.
The CCSD method had to be restricted to the calculation of
the energies of the optimized structures of I and II, their dipole
moments, and their electronic field gradients, which allow the
nuclear quadrupole coupling constants of the 14N nucleus to be
calculated. Bailey’s program AXIS30,37 was used to transform
the dipole moment components from the Gaussian03 standard
orientation to the inertial principal-axis system, and his program
NQC37 was employed to calculate the principal-axis nuclear
quadrupole coupling constants of the 14N nucleus.
The CCSD structures of I and II are listed in Table 1.
Inspection of this table reveals that there are again no unusual
bond lengths and bond angles in the two conformers except the
C3−C1−C2−O4 dihedral angle of I, which is predicted to be
147.2° in this case. Conformer II was once more predicted to
have Cs symmetry.
The rotational constants, the principal-axes dipole moments,
and the nuclear quadrupole coupling constants are displayed in
Table 2. This table also lists the planar moment, Pcc, defined by
Pcc = 1/2(Ia + Ib − Ic), where Ia, Ib, and Ic are the principal
moments of inertia, because its value is a sensitive indicator of
the degree of nonplanarity. The CCSD electronic energy
difference between I and II is 2.11 kJ/mol, again favoring I
(Table 2).
DFT calculations generally require considerably less
computational resources than MP2 or CCSD calculations,
and this method has therefore been used to calculate a series of
parameters that would otherwise require considerably more
extensive calculations. The B3LYP method was used to
calculate structures, energies, dipole moments, harmonic and
anharmonic vibrational frequencies, Watson’s A-reduction
quartic and sextic centrifugal distortion constants,38 and the
vibration−rotation constants (α).39
The B3LYP structures of I and II are shown in Table 1.
Interestingly, the C3−C1−C2−O4 dihedral angle of I was
predicted to be 150° in this case. The B3LYP principal-axes
dipole moments and electronic energy difference of 3.55 kJ/
mol are displayed in Table 2. The energy difference becomes
3.71 kJ/mol after correction for zero-point vibrational energies.
Table 1. CCSD/cc-pVTZ and B3LYP/cc-pVTZ Structures of
Conformers I and II of Cyanoacetaldehyde
method
CCSD
conformer
I
C1−C2
C1−C3
C1−H6
C1−H7
C2−O4
C2−H8
C3−N5
C2−C1−C3
C2−C1−H6
C2−C1−H7
C3−C1−H6
C3−C1−H7
H6−C1−H7
C1−C2−O4
C1−C2−H8
O4−C2−H8
C1−C3−N5
C3−C1−C2−O4
C3−C1−C2−H8
H6−C1−C2−O4
H6−C1−C2−H8
H7−C1−C2−O4
H7−C1−C2−H8
a
B3LYP
II
Bond Length (pm)
152.5
152.0
146.5
146.3
108.8
109.2
109.2
109.2
120.0
119.9
110.1
110.2
115.5
115.4
Angle (deg)
112.0
113.6
109.2
108.5
107.9
108.5
110.8
109.7
108.6
109.7
108.2
106.6
122.2
124.7
115.8
113.6
122.0
121.8
178.2a
178.8b
Dihedral Angle (deg)
−147.2
0.0
34.3
180.0
−24.1
122.3
157.4
−57.7
93.3
−122.3
−85.3
57.7
I
II
153.2
145.6
109.0
109.5
119.9
110.5
114.9
152.5
145.4
109.5
109.5
119.6
110.7
114.9
112.9
108.9
107.2
111.1
109.0
107.5
122.1
115.7
122.2
178.0a
112.9
107.9
107.9
110.0
110.0
105.8
125.3
114.9
121.8
179.1b
−150.1
31.3
−26.1
155.3
89.9
−88.7
0.0
180.0
123.0
−56.9
−123.0
56.9
Bent toward H8. bBent away from O4.
The harmonic and anharmonic vibrational frequencies of I are
listed in Table 6S of the Supporting Information, whereas the
vibrational frequencies of II are listed in Table 8S. The Watson
quartic and sextic A-reduction centrifugal distortion constants38
of conformer II are listed in Table 5. The vibration−rotation
constants are shown in Tables 7S (conformer I) and Table 9S
(II) of the Supporting Information.
The fact that the low-energy forms I and II are both
conformers of cyanoacetaldehyde makes it of interest to explore
the potential energy functions for rotation about the C1−C2
bond, which separates the two forms. Electronic energy
potential functions were therefore calculated at both the
MP2/aug-cc-pVTZ and B3LYP/aug-cc-pVTZ levels of theory
employing the scan option of Gaussian03. The energies were
computed in steps of 10° of the C3−C1−C2−O4 dihedral
angle. All the remaining structural parameters were optimized
for each dihedral angle. Separate calculations of the energies
and optimized structures of the transition states near 60° and at
180° were also performed. The potential functions (red circles,
MP2; blue squares, B3YP) based on the results of these
calculations combined with the above results for I and II are
drawn in Figure 2. The global minimum (conformer I) has a
C3−C1−C2−O4 dihedral angle of 145° (MP2) and 150°
(B3LYP). The second minimum at 0° found in both methods
corresponds to conformer II. The electronic energy differences
between I and II are 2.33 (MP2) and 3.55 kJ/mol (B3LYP).
The energy of the transitions state at 61° are 9.66 (MP2) and
10.59 kJ/mol (B3LYP), respectively.
Most interestingly, both methods predict very low values for
the second transition state at the exact antiperiplanar (180°)
4050
dx.doi.org/10.1021/jp212306z | J. Phys. Chem. A 2012, 116, 4047−4056
The Journal of Physical Chemistry A
Article
vibrational state of conformer I. These energy levels are often
designated with a (+) for the lowest-energy level and a (−) for
the higher-energy level. Moreover, coupling between vibration
and rotation is often observed in such low-barrier cases, often
resulting in complicated rotational spectra that cannot be fitted
satisfactorily to the standard Watson Hamiltonian,38 which
does not take this effect into account. Complicated vibration−
rotation coupling was indeed found to be a prominent
perturbation in the rotational spectrum of I (see below).
Quantum chemical model calculations are often a very useful
aid for the assignment of rotational spectra complicated by
vibration−rotation coupling. Recently, such calculations were
successfully performed for thiophenol and 4-fluorothiophenol,40 which present similar problems as that of I, associated
with the large-amplitude vibration of the thiol groups in these
two compounds. It was therefore decided to undertake a similar
modeling for I, using the same method.40 This procedure40
combines an internal rotation program called SEMIRIG with
the program ASMJ5. The latter computer program fits
vibration−rotation energy levels for J ≤ 7 to rotational and
centrifugal distortion constants as well as Coriolis coupling
constants for two closely spaced vibrational levels.40 The
SEMIRIG program has the internal rotation of one part of the
molecule, the CH2(CN) group in our case, against the other
part, the CHO group, as the only internal degree of freedom
and has no symmetry restriction on the two parts of the
molecule. A detailed account of this program has been
published.41
The internal rotation potential function in SEMIRIG has the
form V(γ) = Σ1/2Vn(1 − cos(nγ)), where γ is the internal
rotation angle. Only the first three coefficients of this
expansion, V1−V3, were employed in the present calculations.
Their values, which are listed in Table 3, were obtained from
the B3LYP potential function shown in Figure 2, making use of
the minima at 0 and 150° and the maximum at 61°. Necessary
Table 2. CCSD/cc-pVTZ and B3LYP/cc-pVTZ
Spectroscopic Constants of Conformers I and II of
Cyanoacetaldehyde
method
conformer
A
B
C
Pcca
μa
μb
μc
μtot
χaa
χbb
χab
ΔE
CCSD
I
B3LYP
II
I
Rotational Constants (MHz)
26 034.6
12 886.2
26 933.0
2584.7
3643.4
2569.6
2438.9
2890.5
2425.1
CCSD Planar Moment (10−20 u m2)
3.86
1.55
4.60
Dipole Momentb (10−30 C m)
3.11
8.98
3.77
5.25
16.95
4.94
4.25
0.0c
3.43
7.44
19.18
7.10
Quadrupole Coupling Constantd (MHz)
−3.600
−2.399
1.239
−0.097
3.036
3.664
Electronic Energy Difference Relative to I (kJ/mol)
0.0e
2.11
0.0f
II
13 111.8
3579.8
2861.0
1.54
8.13
15.51
0.0c
17.51
3.55g
Planar moment defined by Pcc = 1/2(Ia + Ib − Ic), where Ia, Ib, and Ic
are the principal moments of inertia. Conversion factor: 505379.05 ×
10−20 MHz u m2. bOne debye = 3.33564 × 10−30 C m. cFor symmetry
reasons. dThe nuclear quadrupole coupling constants of the nitrogen
nucleus were calculated only at the CCSD level of theory. eCCSD
electronic energy of I: −644897.81 kJ/mol. fB3LYP electronic energy
of I: −646294.48 kJ/mol. gThe energy difference becomes 3.71 kJ/mol
when corrected for zero-point vibrational energies.
a
Table 3. Potential Constants, Mean Rotational Constants,
Energy Difference, and Coupling Constants of Conformer I
of Cyanoacetaldehyde
B3LYPa
Exp.b
c
Potential Constants (kJ/mol)
−7.864
5.141
5.143
Mean Rotational Constantsd (MHz)
Am
26 727.27
26 749.46
Bm
2569.11
2577.69
Cm
2434.63
2425.08
Difference between the Rotational Constants of the Two States (MHz)
ΔA
−429.25
−14.66
ΔB
5.66
3.39
ΔC
7.93
3.02
Planar Moment (10−20 u m2)
Pcc
4.02
3.28
Difference between the (+)- and (−)-States (MHz)
ΔE
68158
58794
Coriolis Coupling Constants (MHz)
Fbc
9.8
0.52
Fac
11.0
0.0c
V1
V2
V3
Figure 2. MP2/aug-cc-pVTZ (red, circles) and B3LYP/aug-cc-pVTZ
(blue, squares) electronic energy potential functions for rotation about
the C1−C2 bond of cyanoacetaldehyde. The C3−C1−C2−O4
dihedral angles are given on the abscissa, and the energies relative to
the global-minimum, conformer I, are given on the ordinate.
conformation. MP2 indicates 0.84, while B3LYP yields 0.48 kJ/
mol for this barrier height.
A double-minimum potential for rotation about the C1−C2
bond with a comparatively small barrier at the exact
antiperiplanar conformation is therefore predicted to exist for
this compound. A barrier of this height should lead to largeamplitude vibrations for the torsion about the C1−C2 bond
and to two closely spaced energy levels for the ground
a
The rotational constants have been obtained in the combined B3LYP
plus SEMIRIG-ASMJ5 calculations. bSee text. cFit 1, Table 4. dFixed.
4051
dx.doi.org/10.1021/jp212306z | J. Phys. Chem. A 2012, 116, 4047−4056
The Journal of Physical Chemistry A
Article
branch series, where K−1 = 2 ← 1. Many of these transitions
were readily assigned because their frequencies were close to
those predicted using the Watson Hamiltonian.38 However,
some of these transitions were definitely shifted from the
frequencies predicted using this Hamiltonian, and Coriolis
coupling between vibration and rotation was suspected of
causing this inconsistency.
Having assigned the bQ-lines, when possible, the assignment
of a-type R-branch lines became the next target. The K−1 ≥ 3
a
R-branch lines occur in pile-ups because I is a very prolate
rotor with Ray’s asymmetry parameter44 κ ≈ −0.99. These
transitions are modulated at comparatively low Stark field
strengths, which makes them easy to assign, and are often good
candidates for the very specific RFMWDR method. The result
of a RFMWDR search is shown in Figure 3. It is seen from this
structural parameters were taken from the B3LYP structure of I
(Table 1). All rotational levels with J ≤ 7 of the (+)- and
(−)-states were fitted precisely by ASMJ5 to obtain effective
rotational constants, plus centrifugal distortion constants of
both states, as well as the Fbc and Fac Coriolis coupling
constants of the two states. The (+)- and (−)-levels were
calculated in this manner to be spaced by 68 158 GHz (about
2.27 cm−1); see Table 3. In the same table, the mean values of
the rotational constants of the (+)- and (−)-states are shown
together with the differences between the upper (−)-state and
the lower (+)-state rotational constants. The B3LYP rotational
constants in Table 3, referring to the ground-vibrational state,
differ a little from their counterparts in Table 2, which refer to
the approximate equilibrium structure. The quartic centrifugal
distortion constants based on the B3LYP plus SEMIRIG
ASMJ5 model calculations were ΔJ = 0.88, ΔJK = −102, ΔK =
3191, δJ = −0.138, and δK = 1.7 kHz for the (+)-state,
remarkably different from the corresponding constants for the
(−)-state, which were 0.63, −68, 2027, −0.088, and −7.2 kHz,
respectively.
Microwave Spectrum and Assignment of I. Conformer
I is predicted in the theoretical calculations above to be
preferred over II by 2−4 kJ/mol and over III by 6.6 kJ/mol. I
and II might be present in so high concentrations that it should
be possible to assign their rotational spectra, while III
presumably has a small Boltzmann factor at room temperature
and has consequently a very weak spectrum, which would
probably be difficult to assign. It was therefore initially assumed
that the spectra of I and II would be responsible for the vast
majority of the observed absorption lines. Both these rotamers
are prolate and have their major dipole moment component
along the b axis (Table 2). The perpendicular b-type spectra of
prolate rotors are rich, and this was indeed observed as the
rotation spectrum was found to have absorption lines occurring
every few megahertz throughout the investigated spectral
ranges.
The K−1 = 1 ← 0 b-type Q-branch transitions of the
spectrum of I were predicted to be among the strongest lines of
the spectrum. Searches were first undertaken in Oslo for these
transitions using the spectroscopic constants of Table 1 to
predict their approximate spectral frequencies. These transitions were soon identified. Most interestingly, these lines
appeared as doublets of apparently equal intensities and were
generally separated by several megahertz. Calculation of the 14N
nucleus quadrupole splittings of these transitions using program
MB0942 and the nuclear quadrupole coupling constants of
Table 2 revealed that the doublets could not at all be explained
by this effect, which would cause relatively small and only partly
resolved splittings of the strongest quadrupole components.
However, this doublet feature and the existence of (+)- and
(−)-states is exactly what the calculations above predict for a
compound possessing a double-minimum potential with a small
barrier maximum at the antiperiplanar position.
The K−1 = 1 ← 0 bQ lines of both the (+)- and the (−)-state
could be fitted reasonably well to Watson’s Hamiltonian38 in
the A-reduction Ir-representation using Sørensen’s program
ROTFIT.43 The fact that the A rotational constant is larger for
the (+)-state than for the (−)-state, whereas the opposite is the
case for B and C, is in accord with the B3LYP calculations
(Table 3). This criterion was used to assign the observed
spectra to the (+)- and (−)-states and not vice versa.
The preliminary spectroscopic constants obtained from these
first Q-branch transitions were used to predict the next Q-
Figure 3. Portion of the RFMWDR spectrum of the J = 12 ← 11
transition obtained employing a radio frequency of 7.50 MHz. The
spectrum shows the J = 123 ← 113 pair of lines of four vibrational
states. The state denoted 1 is assumed to be the (+)-state, and the
state denoted 2 is assumed to be the (−)-state. The states denoted 3
and 4 are other vibrationally excited states, which were not analyzed
further.
figure that the (+)- and the (−)-states have roughly the same
intensity, which is expected for two closely spaced energy levels
with a separation of a few cm−1.
Inclusion of aR-branch transitions together with the bQ-lines
in the least-squares procedure of ROTFIT allowed preliminary
values of spectroscopic constants to be calculated and further btype R-branch lines to be assigned as the next step.
These constants were used to predict the millimeter-wave
spectrum with the Watson Hamiltonian.38 This allowed many
additional transitions to be assigned in Lille. Interestingly,
transitions up to J = 40 and K−1= 7 of the (−)-state could be
fitted satisfactorily to this Hamiltonian. However, many lines of
the (+)-state were definitely perturbed and a more elaborate
Hamiltonian than Watson’s was needed to produce a better fit.
A Hamiltonian, which includes Coriolis coupling similar to
the one used successfully for ethanetellurol,45 seemed to be a
natural choice. The Hamiltonian40 we employed, includes a
coupling between the two states of the type Fbc(Jb̂ Jĉ + Jĉ Jb̂ ). This
procedure gave a very good fit to low-K−1 lines (K−1 < 3), and
it was even possible to include a number of lines, which did not
fit to Watson’s Hamiltonian. These strongly perturbed lines all
belonged to the (+)-state. The spectrum consisting of 199
4052
dx.doi.org/10.1021/jp212306z | J. Phys. Chem. A 2012, 116, 4047−4056
The Journal of Physical Chemistry A
Article
transitions from the 19.5−80 GHz spectral region with K−1 < 3
is listed in Table 10S of the Supporting Information. The
spectroscopic constants obtained from this fit, denoted fit 1, are
listed in the second column of Table 4. The mean values of the
since the only internal degree of freedom in the model
calculations is the torsion about the C1−C2 bond.
More unexpected is maybe the large differences between the
experimental centrifugal distortion constants from the (+) to
the (−) state (Table 4), but since also the model calculations
give large differences between the two states, we suggest that
the differences are caused primarily by the highly anharmonic
potential function with many low-lying torsional states.
The separation between the (+)- and (−)-energy levels was
found to be 58 794(14) MHz [1.96112(5) cm−1] in fit 1 (Table
4), compared to the B3LYP value of 68 158 MHz (Table 3),
which is considered to be satisfactory.
In a second least-squares fit, called fit 2, a total of 413 lines
listed in Table 11S, Supporting Information, were fitted.
Transitions from the (+)-state with K−1 < 3 and lines from the
(−)-state with K−1 < 8 were employed in this fit. The (+)-state
lines are from the 20−80 GHz spectral interval. Several
inconsistencies were observed when (+)-state lines from the
region above 150 GHz were included in the fit, and no such
lines were therefore used to obtain the spectroscopic constants
shown in Table 4. However, many (−)-state transitions from
the high-frequency spectral interval have been used (Table 11S,
Supporting Information). In order to obtain this fit, lines
involving rotational levels of the type J1,J with J = 30, 31, and 32
had to be excluded. Even though three different coupling
operators were used, lines involving these levels were not well
reproduced by the fit. The most disturbing fact is that most of
the lines in the upper state can be fitted pretty well to a usual
Watson Hamiltonian, as shown in fit 3 of Table 4.
Altogether, we consider fit 1 to be best, but the cause of the
problems or inconsistencies encountered is not known. It might
be due to some wrong assignments or the reason could be that
an unknown third state is involved. Finally, and most likely, the
effective Hamiltonian used is not adequate and lacks some kind
of operator or coupling to other states.
The (+)- and (−)-states are separated by 58 794(14) MHz
according to fit 1. The c-type transitions should occur between
these two levels. It is seen from Table 2 that rather significant
values of μc are predicted in the theoretical calculations for
conformer I. Searches were therefore undertaken in an attempt
to assign c-type transitions using the separation by about 58 794
MHz to guide the efforts. However, no c-type lines could be
assigned in this dense spectrum in this manner.
Assignment of the Spectrum of II. The theoretical
calculations above (Figure 1 and Table 2) indicate that rotamer
II should be 2−4 kJ/mol less stable than I. This high-energy
rotamer of cyanoacetaldehyde is predicted to have a
comparatively large μb of about 17 × 10−30 C m (CCSD
value; Table 2), and it should therefore have a relatively strong
b-type spectrum even given an energy difference of a few kJ/
mol. Successful searches for bQ-branch transitions were first
performed in Oslo using the CCSD rotational constants (Table
2) and the B3LYP centrifugal distortion constants (Table 5) to
predict their approximate frequencies. The assignments were
then extended to include additional b-type Q- and R-branch
lines, as well as aR-transitions.
Using the preliminary spectroscopic constants from the Oslo
measurements, the assignments of the millimeter and
submillimeter-wave spectra were fairly easy to perform. The
major difficulty came from the high density of the spectra due
to unassigned lines from excited vibrational states of conformer
I. Hundreds of lines from the first excited state of the lowest
Table 4. Spectroscopic Constantsa of Conformer I of
Cyanoacetaldehyde
fit no.
1
A (MHz)
B (MHz)
C (MHz)
ΔJ (kHz)
ΔJK (kHz)
ΔK (kHz)
δJ (kHz)
δK (kHz)
ΦJ (Hz)
ΦJK (Hz)
ΦKJ (Hz)
ϕJ (Hz)
ϕJK (Hz)
λJK (mHz)
26 756.77(9)
2576.000(4)
2423.565(6)
1.431(17)
−98.9(6)
2635(20)
0.035(6)
−4.9(14)
0.08(2)
20.1(8)
1288(135)
−0.017(6)
−50(4)
23(3)
A (MHz)
B (MHz)
C (MHz)
ΔJ (kHz)
ΔJK (kHz)
ΔK (kHz)
δJ (kHz)
δK (kHz)
ΦJ (Hz)
ΦJK (Hz)
ΦKJ (Hz)
ΦK (Hz)
ϕJ (Hz)
ϕJK (Hz)
ΛJJK (mHz)
ΛJK (mHz)
ΛKK (mHz)
λK (mHz)
Δ (MHz)
Gcb (MHz)
Fbcc (MHz)
hxyJd (kHz)
rmse (kHz)
no. of lines
2
(+)-State
26 756.67(12)
2575.982(5)
2423.562(8)
1.348(21)
−100.8(8)
2628(26)
0.046(8)
−8.5(20)
−0.07(3)
15.3(10)
1067(173)
−0.0002(80)
−38(5)
12(4)
(−)-State
26 742.13(8)
26 741.75(7)
2579.388(4)
2579.3959(20)
2426.588(4)
2426.5975(22)
0.911(15)
0.9291(21)
−35.4(4)
−36.17(8)
394(16)
299(17)
0.1354(2)
0.1426(14)
−22.0(3)
−23.49(23)
−0.013(14)
−0.0002(7)
−0.66(23)
1.04(8)
321(96)
−60(3)
−790(762)
0.0111(15)
−2.12(16)
−0.042(28)
−2.9(9)
195(43)
−0.0028(5)
Coupling Terms
58 794(14)
58 735(17)
0.74(22)
0.5208(15)
1.13(3)
−0.74(4)
0.156
0.204
199
413
3
26 742.006(22)
2579.3873(10)
2426.5878(11)
0.9232(9)
−36.557(29)
361(5)
−0.1349(3)
22.33(12)
−0.0012(3)
0.764(10)
−68.3(16)
1812(201)
−0.0019(1)
1.53(9)
217(23)
0.123
316
a
See text. Uncertainties represent one standard deviation. bOperator:
̂Jc cOperator: (Jb̂ Jĉ + Jĉ Jb̂ ). dOperator: (Jb̂ Jĉ + Jĉ Jb̂ )J2̂ . eRoot-mean-square
deviation.
rotational constants of the (+)- and (−)-states are shown in
Table 3 together with the B3LYP counterparts.
The agreement between the experimental and B3LYP
rotational constants is remarkable, the main difference being
much larger than observed values for ΔA. The theoretical
coupling constants Fbc and Fac are much larger than their
experimental counterparts (Table 3). The quartic centrifugal
distortion constants given in the text above differ much from
the experimental values shown in Table 4. This is expected
4053
dx.doi.org/10.1021/jp212306z | J. Phys. Chem. A 2012, 116, 4047−4056
The Journal of Physical Chemistry A
Article
Table 5. Spectroscopic Constantsa of Conformer II of Cyanoacetaldehyde
vibrational state
ground
first torsion
first bending
equilibriumb
A (MHz)
B (MHz)
C (MHz)
Pcc (10−20 u m2)
ΔJ (kHz)
ΔJK (kHz)
ΔK (kHz)
δJ (kHz)
δK (kHz)
ΦJ (Hz)
ΦJK (Hz)
ΦKJ (Hz)
ΦK (Hz)
ϕJ (Hz)
ϕJK (Hz)
ϕK (Hz)
ΛJ (mHz)
ΛJJK (mHz)
ΛJK (mHz)
ΛKKJ (mHz)
ΛK (mHz)
λJ (mHz)
λJK (mHz)
πJJK (μHz)
rmsd
no. of linese
12 812.11068(26)
3658.693685(84)
2894.587294(87)
1.490958(6)
3.533884(54)
−24.99519(19)
84.6528(17)
1.079689(20)
6.95591(91)
0.011740(15)
−0.003236(95)
−0.54216(30)
2.0026(48)
0.0052897(62)
0.01755(74)
1.07672(85)
−0.0000470(13)
0.0000268(82)
−0.004269(43)
0.02444(22)
−0.0677(41)
−0.00001837(57)
0.00175(19)
−0.000137(16)
0.921
2338
12 790.675(14)
3638.8267(62)
2895.3248(62)
1.9234(1)
3.266(29)
−24.079(60)
79.48(16)
1.0094(15)
1.312(70)
−0.113(13)
−0.080(65)c
12 923.719(19)
3671.230(10)
2894.104(10)
1.0702(1)
3.692(38)
−25.602(79)
93.29(80)
1.1408(17)
12.044(74)
12 886.2
3643.4
2890.5
1.545
3.21
−22.8
84.1
0.950
6.78
0.00427
−0.0170
−0.413
1.81
0.00427
0.0143
0.986
1.709
104
1.474
78
0.452(71)c
a
A-reduction.38 Uncertainties represent one standard deviation. bCCSD rotational constants and B3LYP centrifugal distortion constants. cFurther
centrifugal distortion constants preset at zero. dWeighted root-mean-square deviation. eNumber of lines used in the least-squares fit.
126(25) cm−1 for this vibration, in good agreement with the
B3LYP harmonic value, 134 cm−1.
The vibration−rotation constants,39 α, calculated from α =
X0 − X1, where X0 is a rotational constant of the ground
vibrational state and X1 is a rotational constant of the first
excited state of the normal mode in question, were calculated
from the entries of Table 5 as αA = 21.43, αB = 19.86, and αC =
−0.74 MHz, respectively. This is in fair agreement with the
B3LYP values (Table 9S, Supporting Information), which are
αA = 39.32, αB = 17.09, and αC = −1.36 MHz.
Relative intensity measurements49 yielded 158(30) cm−1 for
the first excited state of the in-symmetry-plane bending
vibration, close to the B3LYP prediction of 156 cm−1. Pcc
decreases for this excited state compared to the ground
vibrational state (Table 5), which is in accord with theory.47,48
The vibration−rotation constants obtained from the values in
Table 5 are αA = −111.61, αB = −12.54, and αC = 0.48 MHz,
compared to the B3LYP values (Table 9S, Supporting
Information) of αA = −125.63, αB = −13.25, and αC = 0.62
MHz, respectively. The agreement is again satisfactory.
Searches for the Tautomer III. This species is
comparatively polar with MP2 dipole moment components of
μa ≈ 10, and μb ≈ 6.5 × 10−30 C m (Table 3S, Supporting
Information). The rather large energy difference of 6.6 kJ/mol
(Figure 1) between this form and conformer I indicates that it
would possess a small Boltzmann factor at room temperature
and, consequently, that the spectrum would be expected to be
very weak. Extensive searches for this spectrum were made
using the MP2 rotational constants and dipole moment
components (Table 3S, Supporting Information) to predict
bending vibration of this conformer could be assigned in the
millimeter-wave spectra.
A total of 2338 lines shown in Table 13S, Supporting
Information, were used to calculate the spectroscopic constants
of the ground vibrational state of rotamer II. These parameters,
which are listed in Table 5, were obtained in a weighted leastsquares fit employing the computer program ASFIT.46 It is seen
from Table 5 that very accurate values have been obtained not
only for the rotational constants but also for the quartic and
sextic centrifugal distortion constants. The experimental quartic
centrifugal distortion constants agree with the B3LYP constants
to within about 10% or better, while much larger discrepancies
exist for some of the sextic constants.
Vibrationally Excited State of II. There exist two lowfrequency vibrational normal modes of this rotamer with
harmonic frequencies at 134 and 156 cm−1, according to the
B3LYP calculations shown in the Supporting Information,
Table 8S. The former of these modes is the torsion about the
C1−C2 bond, whereas the latter is the lowest bending
vibration. A total of 104 transitions were assigned for the
spectrum of the first excited state of the torsion and 78 lines
were assigned for the first excited state of the lowest bending
vibration. These spectra are listed in Tables 14S and 15S of the
Supporting Information, while the spectroscopic constants
obtained in a weighted least-squares calculation using
ROTFIT43 are shown in Table 5.
It is seen from this table that the planar moment, Pcc,
increases for the first excited state of the torsion compared with
the ground vibrational state. This is typical for an out-ofsymmetry-plane vibration.47,48 Relative intensity measurements
performed largely as described by Esbitt and Wilson49 yielded
4054
dx.doi.org/10.1021/jp212306z | J. Phys. Chem. A 2012, 116, 4047−4056
The Journal of Physical Chemistry A
Article
spectra of two rotameric forms of cyanoacetaldehyde produced
by rotation about the C−C bond have been assigned. The
cyano and carbonyl groups forms an unusual angle of 151(3)°
from the synperiplanar (0°) position in I. These groups are
exactly synperiplanar in II. Conformer I is preferred over II by
2.9(8) kJ/mol.
A double minimum potential for rotation about the C−C
bond with a small barrier maximum at the exact anticlinal
(180°) leads to Coriolis perturbations in the rotational
spectrum of conformer I. Selected transitions of the spectrum
of I was fitted to a Hamiltonian allowing for this sort of
interaction and the separation between the two lowest
vibrational states were determined to be 58 794(14) MHz
[1.96112(5) cm−1]. However, attempts to include additional
transitions in the fits using this Hamiltonian failed, and it is
concluded that it lacks interaction terms to account
satisfactorily for all the observed transitions.
The Coriolis perturbation appears to be limited to conformer
I because more than 2000 transitions were assigned and fitted
to the usual Watson Hamiltonian. This large number of
transitions yielded very accurate values not only for the
rotational constants but also for many centrifugal distortion
constants. Two vibrationally excited states were also assigned
for this form, and their vibrational frequencies were determined
by relative intensity measurements.
Theoretical calculations were performed at the B3LYP, MP2,
and CCSD levels of theory using large basis sets to augment the
experimental work. The predictions of these calculations turned
out to be in good agreement with most of the experimental
results, when a comparison could be made.
the spectrum, but no assignments were achieved, presumably
because of its weakness.
Energy Difference between I and II. The energy
difference between I and II was determined from relative
intensity measurements observing the precautions of Esbitt and
Wilson.49 Several selected transitions of both conformers were
used. The internal energy difference between the ground states
of the two forms was calculated as described by Townes and
Schawlow.50 The result was 2.9(8) kJ/mol, with I as the lowestenergy conformer. There are several sources of error involved
in relative intensity measurements,49 and a liberal uncertainty
limit is estimated to be ±0.8 kJ/mol. The energy difference of
2.9(8) kJ/mol should be compared to the theoretical results,
which are 2.3 (MP2), 3.7 (B3LYP), and 2.1 kJ/mol (CCSD),
demonstrating good agreement with experiment.
Structures. It is our experience that CCSD/cc-pVTZ bond
lengths and bond angles of molecules such as cyanoacetaldehyde are very close to the equilibrium values. The bond lengths
and bond angles of conformer I given in Table 1 are therefore
assumed to be our best estimates of these equilibrium
parameters.
A different situation exists for the value of the C3−C1−C2−
O4 dihedral angle, which is of major interest in the case of
conformer I. Figure 1 indicates that the potential curve is rather
flat around 150°. The accurate value of the true minimum of
this curve can therefore be difficult to obtain in the theoretical
calculations.
The C3−C1−C2−O4 dihedral angle is very sensitive to the
value of the planar moment, Pcc. The experimental value is 3.28
× 10−20 u m2 (Table 3), compared to 4.02 (B3LYP; Table 3),
4.19 (MP2, calculated from the entries in Table 1S, Supporting
Information), and 3.86 (CCSD; Table 2). All three theoretical
values are therefore too large, which indicates that the
theoretical values of the C3−C1−C2−O4 dihedral angle,
which were 150 (B3LYP), 145 (MP2), and 147° (CCSD), are
slightly too small.
An improved value of the C3−C1−C2−O4 dihedral angle
was obtained by rotating about the C1−C2 bond keeping the
bond length and bond angles fixed at the CCSD values shown
in Table 1. A value of 151° for the C3−C1−C2−O4 dihedral
angle yielded Pcc = 3.40 × 10−20 u m2, in good agreement with
the experimental value of 3.28 × 10−20 u m2 (Table 3). The
rotational constants obtained in this manner are A = 26 764.2,
B = 2574.0, and C = 2424.8 MHz, which compares very well
with the experimental results shown in Table 3. The
uncertainty limit of the C3−C1−C2−O4 dihedral angle of
conformer I is estimated to be ±3°.
The situation for conformer II is simpler than that for I.
Inspection of Table 5 reveals that there is very good agreement
between the ground-state rotational constants (column 2) and
the CCSD rotational constants (column 5, Table 2) calculated
from the structure in Table 1. The CCSD structure is an
approximation of the equilibrium structure, while the
experimental rotational constants reflect the effective structure
(r0 structure), which is defined differently. Nevertheless, the
good agreement between the experimental and CCSD
rotational constants indicates that the CCSD structure is
indeed close to the equilibrium structure.
■
ASSOCIATED CONTENT
S Supporting Information
*
Results of the theoretical calculations and the microwave
spectra. This material is available free of charge via the Internet
at http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*Tel: +47 2285 5674. Fax: +47 2285 5441. E-mail: harald.
mollendal@kjemi.uio.no.
Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS
We thank Anne Horn for her skillful assistance. The Research
Council of Norway (Program for Supercomputing) is thanked
for a grant of computer time. J.-C.G., L.M., and R.A.M. thank
the program “Physique et Chimie du Milieu Interstellaire”, and
J.-C.G. thanks “Environnements Planétaires et Origine de la
Vie” (INSU-CNRS) and the Centre National d’Etudes
Spatiales (CNES) for financial support.
■
REFERENCES
(1) Luna, A.; Mó, O.; Yáñez, M.; Guillemin, J.-C.; Gal, J.-F.; Maria,
P.-C. Int. J. Mass Spectrom. 2007, 267, 125−133.
(2) Ferris, J. P.; Sanchez, R. A.; Orgel, L. E. J. Mol. Biol. 1968, 33,
693−704.
(3) Claisen, L. Ber. Dtsch. Chem. Ges. 1903, 36, 3664−3673.
(4) Ben, C. A.; Chuche, J.; Manisse, N.; Pommelet, J. C.; Netsch, K.
P.; Lorencak, P.; Wentrup, C. J. Org. Chem. 1991, 56, 970−975.
(5) Jachak, M.; Mittelbach, M.; Kriessmann, U.; Junek, H. Synthesis
1992, 275−276.
■
CONCLUSIONS
The present study has shown that cyanoacetaldehyde
(NCCH2CHO) is strongly preferred over its tautomer
cyanovinylalcohol (NCCHCHOH) in the gas phase. The
4055
dx.doi.org/10.1021/jp212306z | J. Phys. Chem. A 2012, 116, 4047−4056
The Journal of Physical Chemistry A
Article
(6) Frazza, E. J.; Rapoport, L. β-Cyanovinyl Thio Ethers of Thiazole
Compounds, US Patent 3271408, 1966.
(7) Cole, G. C.; Møllendal, H.; Khater, B.; Guillemin, J.-C. J. Phys.
Chem. A 2007, 111, 1259−1264.
(8) Guillemin, J.-C.; Breneman, C. M.; Joseph, J. C.; Ferris, J. P.
Chem.Eur. J. 1998, 4, 1074−1082.
(9) Xiang, Y.-B.; Drenkard, S.; Baumann, K.; Hickey, D.;
Eschenmoser, A. Helv. Chim. Acta 1994, 77, 2209−2250.
(10) Askeland, E.; Møllendal, H.; Uggerud, E.; Guillemin, J.-C.;
Aviles Moreno, J.-R.; Demaison, J.; Huet, T. R. J. Phys. Chem. A 2006,
110, 12572−12584.
(11) Dickens, J. E.; Irvine, W. M; Nummelin, A.; Møllendal, H.;
Saito, S.; Thorwirth, S.; Hjalmarson, A.; Ohishi, M. Spectrochim. Acta,
Part A 2001, 57A, 643−660.
(12) The Cologne Database for Molecular Spectroscopy. http://
www.astro.uni-koeln.de/cdms/molecules.
(13) Sanchez, R. A.; Ferris, J. P.; Orgel, L. E. Science 1966, 154, 784−
785.
(14) Bockelée-Morvan, D.; Lis, D. C.; Wink, J. E.; Despois, D.;
Crovisier, J.; Bachiller, R.; Benford, D. J.; Biver, N.; Colom, P.; Davies.;
et al. Astron. Astrophys. 2000, 353, 1101−1114.
(15) Coustenis, A.; Schmitt, B.; Khanna, R. K.; Trotta, F. Planet.
Space Sci. 1999, 47, 1305−1329.
(16) Horn, A.; Møllendal, H.; Guillemin, J.-C. J. Phys. Chem. A 2008,
112, 11009−11016.
(17) Kawaguchi, K.; Kasai, Y.; Ishikawa, S.; Ohishi, M.; Kaifu, N.;
Amano, T. Astrophys. J. 1994, 420, L95−97.
(18) Stoks, P. G.; Schwartz, A. W. Nature 1979, 282, 709−710.
(19) Ferris, J. P.; Zamek, O. S.; Altbuch, A. M.; Freiman, H. J. Mol.
Evol. 1974, 3, 301−309.
(20) Robertson, M. P.; Miller, S. L. Nature 1995, 375, 772−774.
(21) Robertson, M. P.; Levy, M.; Miller, S. L. J. Mol. Evol. 1996, 43,
543−550.
(22) Nelson, K. E.; Robertson, M. P.; Levy, M.; Miller, S. L. Origins
Life Evol. Biospheres 2001, 31, 221−229.
(23) Cleaves, H. J., II; Nelson, K. E.; Miller, S. L. Naturwissenschaften
2006, 93, 228−231.
(24) Møllendal, H.; Leonov, A.; de Meijere, A. J. Phys. Chem. A 2005,
109, 6344−6350.
(25) Møllendal, H.; Cole, G. C.; Guillemin, J.-C. J. Phys. Chem. A
2006, 110, 921−925.
(26) Samdal, S.; Møllendal, H.; Hnyk, D.; Holub, J. J. Phys. Chem. A
2011, 115, 3380−3385.
(27) Wodarczyk, F. J.; Wilson, E. B., Jr. J. Mol. Spectrosc. 1971, 37,
445−463.
(28) Motiyenko, R. A.; Margulès, L.; Alekseev, E. A.; Guillemin, J.-C.;
Demaison, J. J. Mol. Spectrosc. 2010, 264, 94−99.
(29) Tudorie, M.; Coudert, L. H.; Huet, T. R.; Jegouso, D.; Sedes, G.
J. Chem. Phys. 2011, 134, 074314/1−074314/9.
(30) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao,
O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J.
B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R.
E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.;
Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J.
J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.;
Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman,
J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.;
Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.;
Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen,
W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision
B.03; Gaussian, Inc.: Wallingford, CT, 2003.
(31) Becke, A. D. Phys. Rev. A 1988, 38, 3098−3100.
(32) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785−789.
(33) Møller, C.; Plesset, M. S. Phys. Rev. 1934, 46, 618−622.
(34) Purvis, G. D., III; Bartlett, R. J. J. Chem. Phys. 1982, 76, 1910−
1918.
(35) Scuseria, G. E.; Janssen, C. L.; Schaefer, H. F., III. J. Chem. Phys.
1988, 89, 7382−7387.
(36) Peterson, K. A.; Dunning, T. H., Jr. J. Chem. Phys. 2002, 117,
10548−10560.
(37) Bailey, W. C. Calculation of Nuclear Quadrupole Coupling
Constants in Gaseous State Molecules, 2011. http://web.mac.com/
wcbailey/nqcc.
(38) Watson, J. K. G. Vibrational Spectra and Structure; Elsevier:
Amsterdam, The Netherlands, 1977; Vol. 6, pp 1−89.
(39) Gordy, W.; Cook, R. L. Microwave Molecular Spectra. In
Techniques of Chemistry; John Wiley & Sons: New York, 1984; Vol.
XVII.
(40) Larsen, N. W.; Schulz, L. J. Mol. Struct. 2009, 920, 30−39.
(41) Larsen, N. W.; Steinarsson, T. J. Mol. Spectrosc. 1987, 123, 405−
25.
(42) Marstokk, K.-M.; Møllendal, H. J. Mol. Struct. 1969, 4, 470−472.
(43) Sørensen, G. O. J. Mol. Spectrosc. 1967, 22, 325−346.
(44) Ray, B. S. Z. Phys. 1932, 78, 74−91.
(45) Motiyenko, R. A.; Margulès, L.; Goubet, M.; Møllendal, H.;
Konovalov, A.; Guillemin, J.-C. J. Phys. Chem. A 2010, 114, 2794−
2798.
(46) Kisiel, Z. Programs for Rotational Spectroscopy. http://info.
ifpan.edu.pl/∼kisiel/prospe.htm
(47) Herschbach, D. R.; Laurie, V. W. J. Chem. Phys. 1962, 37, 1668−
1686.
(48) Herschbach, D. R.; Laurie, V. W. J. Chem. Phys. 1964, 40, 3142−
3153.
(49) Esbitt, A. S.; Wilson, E. B. Rev. Sci. Instrum. 1963, 34, 901−907.
(50) Townes, C. H.; Schawlow, A. L. Microwave Spectroscopy;
McGraw-Hill: New York, 1955.
4056
dx.doi.org/10.1021/jp212306z | J. Phys. Chem. A 2012, 116, 4047−4056